COMMENTARY

CKD and Low Magnesium: Pay Attention

Lynda Szczech, MD

Disclosures

February 06, 2013

In This Article

Description of the Study

The analysis of the Framingham Offspring Study under discussion was to determine if a risk for atrial fibrillation was associated with low serum magnesium among people in the community. This study enrolled individuals in 1971 who were the children or the spouses of the children of the original Framingham cohort. The analysis examined correlated serum magnesium levels in 3530 of these individuals with the risk for new atrial fibrillation or atrial flutter seen on subsequent ECGs obtained for the study. When examined in groups based on quartiles of serum magnesium, those in the lowest quartile of serum magnesium were slightly different from those in other quartiles: they were composed of slightly greater proportions of individuals on antihypertensives (12 vs 7%-8%) and individuals with diabetes (8 vs 2%-4%). When various baseline factors were correlated with serum magnesium levels, as would be expected, serum creatinine correlated directly. As serum creatinine increased, so did serum magnesium (P < .0002). A significant and direct relationship was also demonstrated between serum magnesium level and levels of other factors, including serum total/high-density lipoprotein cholesterol ratio, albumin, hemoglobin, calcium, and potassium (all P < .0001).

When adjusted for differences in age and sex among quartiles, the lowest magnesium group had the highest incidence of new atrial fibrillation events (9.4 episodes per 1000 person-years). The rates for the other 3 quartiles were quite similar (6.9, 7.1, and 6.3 events per 1000 person-years for the second, third, and fourth quartiles, respectively). In multivariable models controlled for the multiple potential differences including glomerular filtration rate, those with the lowest levels of magnesium continued to be at highest risk for the development of atrial fibrillation (hazard ratio, 1.48; 95% confidence interval, 1.02- 2.14; P = .04).

processing....